A Multi-Objective Carnivorous Plant Algorithm for Solving Constrained Multi-Objective Optimization Problems.

Area of Science:

• Optimization
• Computational Intelligence
• Algorithm Design

Background:

• Solving constrained multi-objective optimization problems (CMOPs) presents significant challenges due to simultaneous constraint satisfaction and multiple objective management.
• Existing algorithms often struggle to balance exploration and exploitation, leading to suboptimal solutions or slow convergence.

Purpose of the Study:

• To propose a novel optimization algorithm that effectively handles constraints and multiple objectives.
• To enhance the efficiency, global convergence, and precision of optimization algorithms for CMOPs.

Main Methods:

• A Carnivorous Plant Algorithm (CPA) is developed, integrating multi-population and multi-stage strategies.
• The ϵ-constraint handling method is employed with stage-specific ϵ value adjustments.
• Cross-pollination inspired by carnivorous plant behavior, a quasi-reflection learning mechanism, and quadratic interpolation for reproduction are incorporated to improve search capabilities.

Main Results:

• The proposed CPA demonstrates a balance between exploration and exploitation, accelerating convergence.
• Enhanced optimization efficiency and global convergence ability are achieved through novel learning and reproduction mechanisms.
• The algorithm successfully escapes local optima, improving precision and convergence speed.

Conclusions:

• The novel CPA offers a competitive and effective approach for tackling CMOPs.
• Experimental results on diverse test suites confirm its superior performance compared to state-of-the-art algorithms.
• The integration of biological inspiration significantly enhances optimization capabilities for complex problems.